STATISTICALLY OPTIMAL DATA PROCESSING ALGORITHMS FOR SMALL-APERTURE SEISMIC ARRAYS: TESTING ON RECORDED EARTHQUAKES

A. F. Kushnir, V. M. Lapshin, J. Fyen, and T. Kvarna

Abstract

This paper describes a package of statistically optimal data processing algorithms and software application programs for small-aperture seismic arrays, giving some test results using records of seismic events at Scandinavian small-aperture seismic arrays. The package is designed for detecting and identifying noise-contaminated low amplitude seismic signals excited by low magnitude teleseismic or regional events. It can be used to automate the compilation of catalogs of regional and local earthquakes and for monitoring underground nuclear and chemical explosions. The package has an adaptive structure and includes a subsystem for statistical estimation of the matrix spectral density of the seismic noise recorded at the array. The use of this spectral density significantly enhances the signal/noise ratio by suppressing the regional noise component. Testing the package at the NORSAR regional seismicity monitoring system has shown that adaptive statistical data processing when applied to Scandinavian microseisms enhances the power signal/noise ratio by factors of 10 to 30 and significantly improves the accuracy of automatic determinations of azimuths, apparent velocities, and onset times for low amplitude seismic phases from regional events.

Back to
Computational Seismology, Vol. 3.